In its third report, Accounting for Social Risk Factors in Medicare Payment: Criteria, Factors, and Methods (NASEM, 2016), the committee identified social risk factors that the Centers for Medicare & Medicaid Services (CMS) could include in quality measurement and payment, criteria to identify these factors, and methods to do so. For CMS to account for social risk factors in Medicare quality measurement and payment programs using these approaches, it is logical that it must first have accurate data on the social risk factors of Medicare beneficiaries. This chapter describes three broad categories of data sources for these social risk factors: (1) data CMS already possesses or could collect; (2) data that providers (including hospitals, health plans, provider groups, and others) could report to CMS; and (3) alternative government data sources. The chapter also presents general advantages of each potential source as well as barriers to collecting accurate data through, and using data from, these sources.
Patients are the underlying source of most social risk factor data. This is also true of most clinical data. Clinicians make assessments and diagnoses based on how patients present—e.g., their complaints, symptoms, and test results. Providers then systematically maintain and report clinical data in the form of diagnostic and clinical assessments. Most social risk factors are collected directly from patients who report their income, race, ethnicity, preferred language, etc. to CMS, health care providers, and other government agencies. Moreover, for some social risk factors like race, ethnicity, and gender, it is important for patients to self-identify. However, CMS, health care providers and health plans, and government agencies collect and maintain this information and, more importantly, standardize, assess, interpret, and report this information in a valid, consistent, and reliable way.
In the future, new, better, and easier methods of data collection could emerge (e.g., methods that are more accurate, less burdensome, or less costly). For example, health technologies such as smartphone applications and wearable devices that could collect health and social risk factor data are rapidly developing and it is feasible that Medicare beneficiaries could directly report social risk factor data to CMS in the future. Indeed, as these new methods emerge, an ideal system would be responsive to evolving data availability and could adapt to use new data sources. However, at this time and likely in the near term over which the committee expects the Office of the Assistant Secretary of Planning and Evaluation to begin preliminary analyses and CMS to begin accounting for social risk factors in Medicare payment, it is unlikely that technologies and interoperable systems will be available for patients to directly, systematically, and securely submit social risk factor data to CMS for use in Medicare payment. Thus, although
patients and enrollees underlie each of the three categories of data sources described above, they are not called out as a separate and unique source.
Finally, although social risk factor data could also be obtained from private data sources, because these sources and their data collection methods are not fully transparent and because CMS would have to purchase these data at unknown cost, the committee deemed use of such private data as out of scope.
CMS possesses a variety of data sources, some of which include data on social risk factors of Medicare beneficiaries. Existing sources of social risk factor data include administrative records and surveys of enrollees and patients. Administrative records include Medicare beneficiary enrollment records as well as claims data. These sources include limited information on social risk factors, such as beneficiaries’ race and ethnicity (ResDAC, 2016a). Enrollment information on Medicare beneficiaries includes the basis of a beneficiary’s entitlement, which plans beneficiaries are enrolled in (Parts A, B, C, D, or alternative payment models), as well as Medicaid enrollment for those who are dually enrolled in Medicare and Medicaid (ResDAC, 2016a). Administrative records also include basic demographic information and vital statistics, as well as clinical information and data on beneficiaries’ health care use and expenditures based on claims data that providers submit to CMS for payment (ResDAC, 2016a).
Survey data from CMS refer to data derived from one of the surveys of Medicare beneficiaries that CMS routinely conducts. These include the Consumer Assessment of Healthcare Providers and Systems (CAHPS) family of surveys, the Health Outcomes Survey (HOS), and the Medicare Current Beneficiary Survey (MCBS) (ResDAC, 2016b). CAHPS surveys aim to assess patient experiences of care from a variety of care settings—hospital, health plan, clinicians and groups, home health, hospice, and so on (AHRQ, 2016; CMS, 2016a). The Medicare HOS assesses patient-reported health outcomes, including physical functioning and mental health outcomes (Haffer and Bowen, 2004; Medicare Health Outcomes Survey, 2016). The MCBS aims to assess beneficiaries’ access to, satisfaction with, and usual sources of care, as well as their expenditures and sources of payment for all health care services used, including those not covered by Medicare (CMS, 2016c,d). These surveys, especially the CAHPS surveys, include limited data on social risk factors, such as information on race and ethnicity, language, and education.
CMS could also collect new data on social risk factors. It could do so by adding items to existing sources, such as enrollment forms or survey questionnaires. In addition, CMS could collect social risk factor data through new methods or sources, such as through a new survey or administrative form. CMS could implement this for all new beneficiaries going forward, for example, at enrollment as a condition of receiving benefits. However, this would not capture social risk factor data for existing beneficiaries. Thus, to ensure accurate data on all beneficiaries, CMS could also conduct a one-time, universal survey of all currently enrolled Medicare beneficiaries.
Using CMS data has several advantages. The primary advantage of using existing sources of data that CMS already possesses is precisely that CMS has access to and maintains accurate data it already collects using standardized measures and validated, reliable methods, and which it could apply to performance measurement and payment programs. Additionally, if CMS were to collect new social risk factor data for inclusion in Medicare quality measurement and payment
programs, it could design measures and data collection methodologies to ensure collection of accurate data that meet the needs of the intended method to account for those social risk factors in Medicare quality measurement and payment programs. At the same time, such new data collection on the social risk factors also need not be restricted to Medicare quality measurement and payment applications. CMS could also use these data for other purposes, including research and quality improvement. Finally, if CMS were to collect new data themselves, it also would not be subject to the potentially substantial barriers of collaborating with other federal government agencies. (These barriers are discussed in more detail in the section on other government data sources.)
At the same time, using CMS data on social risk factors in quality measurement and payment programs is not without challenges. In particular, although CMS may currently collect and maintain some existing sources of data on social risk factors, because these data are designed and used for purposes that are not quality or performance measurement and payment, these data may not be immediately usable for such application. In particular, data on the same social risk factor across different data sets may require measurement standardization in order to be useful for inclusion in performance measurement and payment. For example, although CMS may have data on the race and ethnicity or preferred language of Medicare beneficiaries from several sources, how such data are measured and/or collected for administrative purposes may differ from how they are measured and/or collected through surveys. Additionally, some data, especially those derived from sample surveys, may not be sufficient for certain methods of accounting for social risk factors. A particular concern is small sample size. For example, CMS would need relatively large sample sizes for some methods of accounting for social risk factors, and this may be larger than what is currently collected through any existing survey. Relatedly, even if sufficient samples are available to account for social risk factors in measuring some outcomes, data on social risk factors from one source may not generalize or be able to be applied to other outcome measures from another source. In regards to new data collection, doing so would require clearance of new items to survey questionnaires or administrative form from the White House Office of Management and Budget, which is especially concerned about collection burden, and such clearance processes could be a barrier to collecting new data. Additionally, any new collection of data from all new or existing Medicare beneficiaries would require substantial cost for which there are likely to be limited resources.
Data sources from providers include data from electronic health records (EHRs) and administrative data that providers report or could report to CMS. EHRs comprise the software providers use to collect, store, and manage patient health records as well as the databases that hold this information (IOM, 2014). EHR data sometimes (and henceforth in this report) refer to the information rather than the entire information technology system (IOM, 2014). Most EHRs capture some basic information on social risk factors, such as race and ethnicity, and EHRs are beginning to capture more robust social risk factor data. Some more comprehensive EHR systems may include or link to more data on social risk factors, such as language preferences or capabilities, education, housing, and community context (Gottlieb et al., 2015; ONC, n.d.).
The Office of the National Coordinator for Health Information Technology (ONC) is the office responsible for supporting and encouraging EHR adoption and health information exchange in the Department of Health and Human Services (HHS). To date, ONC has included
some social risk factors in the regulations it puts forth for the CMS meaningful use incentive programs. Meaningful use in health information technology refers to the use of EHR data for specific objectives, such as quality improvement, care coordination, and improving public and population health (CMS, 2016b; IOM, 2014; ONC, 2014b). CMS meaningful use incentive programs provide bonus payments to providers who demonstrate that their EHRs achieve certain meaningful use objectives. These programs have been implemented in stages since 2011. The Stage 2 Meaningful Use regulations published in 2012 for the incentive program beginning in 2014 require EHRs to have the capacity to include race and ethnicity and preferred language in the objective to record demographics (CMS, 2012). In the Stage 3 Final Rule published in 2015 for programs beginning in 2017, ONC added collection of sexual orientation and gender identity (CMS, 2015). Importantly, achieving meaningful use under these standards does not require providers to collect this information, only that a provider’s EHR system has the capacity to do so (CMS, 2015). Nonetheless, meaningful use regulations and related incentive payments are powerful tools to encourage adoption of social risk factor data in EHRs.
In 2014, the Institute of Medicine (IOM) published a report recommending social and behavioral domains and measures for ONC to consider including in its meaningful use regulations. Although the purpose of that report was to identify social and behavioral domains that should be captured in EHRs to enhance patient care by capturing information important to providers in providing health care, there is some overlap between the social risk factors listed in this earlier IOM report and those identified in the committee’s third report. Moreover, although the tasks for the two committees and the resulting two reports diverge, application of EHR data in Medicare performance measurement and payment can be considered another form of meaningful use and such application provides additional rationale for incentivizing widespread adoption of standardized collection and reporting of data from EHRs to CMS, including social risk factor data.
Administrative data include data captured through patient enrollment forms and claims data and may also include limited social risk factor data. For example, many health plans collect language data in order to provide appropriately tailored health care information and services to enrollees (Lawson et al., 2011; Nerenz et al., 2013a,b), and these data could be reported to CMS for use in performance measurement and payment. Such data could be attached to claims data that providers already submit to CMS using standardized reporting processes and systems for payment.
A primary advantage of using data on Medicare beneficiaries’ social risk factors that providers or health plans collect is that some information on social risk factors may be clinically useful to enhance the care or services the providers and plans provide. In addition, CMS already has a reporting infrastructure for claims and performance reporting with standardized reporting requirements, processes, and systems that it could build on.
Despite these advantages, a principal barrier to using data from providers is the need for standardized measurement and reporting to CMS, regardless of whether the data come from EHRs or other electronic systems. Although CMS has infrastructure for both performance and claims reporting that it could enhance to include reporting of social risk factors, because only limited social risk factor data are currently collected through EHRs, CMS would still need to identify or develop and validate measurement standards for collection of new social risk factors. In addition, data can be added to EHR and other electronic systems through different modes of collection. Clinicians and nonclinicians can collect data through clinical discussions and interviews during an office visit, patients can enter information directly through patient portals or
electronic surveys, and data can be collected through paper forms and entered into electronic systems manually (IOM, 2014). Because these different modes of collection may affect the accuracy and consistency of the data collected, validated data collection methods are needed.
With respect to EHRs in particular, even if measurement and data collection standards are met where they exist, EHR systems lack interoperability, which in turn restricts health information exchange (HIE). HIE is the ability of health care professionals and patients to share patient health records securely and appropriately (ONC, 2014a); whereas, interoperability refers to the architecture and standards that enable HIE across different EHR systems (ONC, 2013). CMS promotes standardized data measurement and collection to promote interoperability and facilitates HIE for EHRs through such programs as the CMS meaningful use incentive programs and regional extension centers that provide technical advice on EHR implementation. However, there are hundreds of vendors of ONC-certified EHRs whose products differ (ONC, 2016e).
Several additional barriers to meaningful use more broadly present additional challenges for the use of EHRs as a source of social risk factor data for use in Medicare quality measurement and payment. Although the number of providers with basic and comprehensive EHRs has risen substantially since 2009 when the ONC was established, in 2015, while nearly all non-federal acute care hospitals used certified EHRs (ONC, 2016b), nearly one-quarter of primary and specialty physician practices did not demonstrate meaningful use of a certified EHR (ONC, 2016c,d). Moreover, evidence suggests that providers disproportionately serving socially at-risk populations such as safety-net hospitals, critical access hospitals, and community health centers are less likely to have either basic or comprehensive EHRs (Adler-Milstein et al., 2014, 2015; DesRoches et al., 2012, 2013a,b; ONC, 2016a,b; Shields et al., 2007). This may be especially challenging for using social risk factor data derived from EHRs in Medicare quality measurement and payment programs because the providers whose performance scores and financial incentives are likely to be most affected by accounting for social risk factors in Medicare quality measurement and payment are precisely those who are less likely to have EHR systems with high functionality.
Second, collecting social risk factor data through EHRs could increase burdens on individual providers and health care organizations, as well as on patients. Adding social risk factors to EHRs may require software upgrades or additional programming; modifying workflows of the clinical team to collect, enter, and manage social risk factor data in the EHR; educating providers on data collection methods to ensure accurate data; ensuring data storage systems and methods to share social risk factor data with other providers and administrators or researchers are secure; and, in some cases, intervening on or otherwise addressing social risk factors through tailored care approaches or referring patients to social service or public health agencies or community organizations that can address unmet social needs (IOM, 2014). Each of these tasks is costly and time consuming.
Burdens on patients and enrollees pertain to the ability of patients to recall information about their social risks as well as privacy and security. With respect to the former, patients and enrollees may not know or be willing to share data on certain social risk factors that are sensitive in nature. Concerns about why clinicians or plans are asking about social risk factors like education, income, or nativity and how such data may be used relate to concerns about the privacy and security of patient health information, especially when shared with other providers and with researchers and administrators for nonclinical uses. The Privacy and Security Rules of the Health Insurance Portability and Accountability Act of 1996 (HIPAA) establishes standards for the use and disclosure of identifiable health information as well as security safeguards to
protect electronic identifiable health information.1 In addition to federal regulations, states and localities also have privacy and security regulations governing the use of health information, which may include social risk factor data (IOM, 2014). Although such privacy laws are important for protecting patient privacy and security, they can nevertheless be barriers to using patient health information for research or administrative purposes. The burden of collecting data on social risk factors in EHRs on patients, providers, and health care organizations, including the concerns described and extensive discussions of privacy and security issues as well as mitigation strategies are discussed in detail in the IOM’s 2014 report as well as the 2012 report on health information technology and patient safety, and the committee points the interested reader to these reports for a more comprehensive discussion (IOM, 2012, 2014).
Alternative government data sources in this report refer to administrative data and national surveys that federal agencies other than CMS and state agencies oversee and maintain and that could be linked to Medicare beneficiary data or that CMS could otherwise use. (Note this includes other agencies within HHS.) These data include data that could be linked to Medicare beneficiary data at the individual level, area-level data that could be used to describe a Medicare beneficiary’s residential environment or serve as a proxy for individual effects, and data that could help CMS to determine how to elicit information on social risk factors from Medicare beneficiaries. The primary advantage of using administrative and survey data from other agencies is that these data sources contain substantial information on social risk factors, and data from these sources are collected using standardized and validated measures and methodologies. However, barriers to linking such data to Medicare data can be substantial. First and foremost, laws and regulations relating to the privacy and security of such data, particularly federally funded data, may restrict data sharing (IOM, 2014). Additionally, as described above, even if data can be shared, it may require substantial effort and/or cost to ensure that data can be linked at the appropriate level. Small sample sizes in surveys may be of particular concern. For example, sample sizes for small geographic areas are small, and data may need to be pooled across years. Furthermore, because data from alternative government sources are not intended for use in Medicare quality measurement and payment applications, the social risk factor variables available from these sources may not best capture the relevant latent constructs. For example, the National Health and Nutrition Examination Study (NHANES) captures sexual orientation data, but focuses on sexual behavior; whereas, the aspect most relevant to Medicare performance indicators may be sexual identity. Barriers specific to particular data sources are discussed in more detail in the following sections.
Data from the Social Security Administration
The Social Security Administration (SSA) may be the most useful source of administrative data on social risk factors outside of CMS that could be linked to Medicare beneficiary data at the individual level. The SSA maintains many different data sets, but the four most commonly used are the Master Beneficiary Record, Master Earnings File, Numident file, and Supplemental Security Record (McNabb et al., 2009). These records include data on
1 45 CFR Part 160 and Subparts A, C, and E of Part 164.
demographics, vital statistics, lifetime earnings (i.e., income), and information related to eligibility for social security needs-based benefits, such as disabling conditions and living arrangements (McNabb et al., 2009). Although these data are available and could be linked to Medicare beneficiary data, doing so will likely require substantial effort and cost.
Data from the American Community Survey
The American Community Survey (ACS) may be a particularly useful source of area-level social risk factor data that could be used to assess genuine area-level effects or serve as proxies for individual-level effects (U.S. Census Bureau, 2016). The ACS is a continuous nationwide survey administered by the Census Bureau that provides a wide range of social risk factor data including demographic, housing, social, and economic data on local communities (U.S. Census Bureau, 2013). It replaced the Census long form beginning in 2000, with full implementation in 2005. The sample size has increased from 2.9 million housing units in 2005 to 3.3 million housing units in 2014 (U.S. Census Bureau, 2013, 2015). Data from the ACS may be particularly useful for Medicare quality measurement and payment applications, because it provides area-level data at relatively small geographies—census tracts and block groups. However, because samples sizes are small, only 5-year estimates are available for these small geographies (U.S. Census Bureau, 2016). Moreover, these estimates are derived from all persons residing in the households sampled, not only the population of older adults. Therefore, they may be appropriate measures of genuine area-level effects but less precise as proxies for individual-level effects. However, deriving census-tract level variables from the ACS exclusively from the population of older adults, under the assumption that these would serve as better proxies for individual-level effects, is likely to be unwise. The small sample sizes of older adults in most census tracts would result in considerable imprecision.
Data from Other National Surveys
Other national surveys could be useful to CMS to determine how best to elicit information from Medicare beneficiaries on their social risk factors, because they capture substantial information on social risk factors and may offer guidance on the potential measurement strategies—both new or alternative ways—for many of the social risk factors that CMS could include in Medicare quality measurement and payment. For example, the design of these surveys includes standardized and validated measures and data collection methods to which CMS could refer when developing and refining its own measures and strategies to collect social risk factor data. However, because sample sizes of older adults in these national surveys are small, data from these surveys is unlikely to be useful to link to Medicare beneficiary data at the individual-level for use in Medicare quality measurement and payment. At the same time, where social risk factor data from national surveys can be linked to individual-level Medicare beneficiary data in some limited capacity, in some cases, these national surveys could serve as test beds for CMS to assess the value-added quality of more complex measures. For example, CMS could assess how much additional explanatory power wealth might have above and beyond other measures of SEP, such as education and income, with regard to performance indicators used in value-based payment.
National surveys that collect data on social risk factors and which may be useful to CMS are the Health and Retirement Survey (HRS), National Health & Aging Trends Study (NHATS), NHANES, National Health Interview Survey (NHIS), and the National Survey of Family Growth
(NSFG). HRS, sponsored by the National Institute on Aging (NIA) and the SSA and administered by the University of Michigan, assesses health and economic well-being among more than 37,000 adults age 50 living in 23,000 households (NIA et al., 2007; Sonnega et al., 2014). NHATS, sponsored by NIA and conducted by the Johns Hopkins Bloomberg School of Public Health with data collection by Westat, assesses late life functioning among more than 8,0000 adults age 65 and older (NHATS, 2016). Because both HRS and NHATS are surveys of older adults including Medicare beneficiaries, some Medicare data are already linked to data from the HRS and NHATS (ResDAC, n.d.-a, n.d.-b). NHIS, NHANES, and NSFG are health-related surveys overseen by the National Center for Health Statistics of the Centers for Disease Control and Prevention. NHIS is a continuous household survey of adults that assesses physical and mental health status, chronic disease, health insurance and access to health care services, health behaviors (e.g., smoking, alcohol use, physical activity, immunizations), and limitations on activity or functioning (CDC, 2015b). The NHIS sample size for surveys beginning in 2011 is expected to be 87,500 persons from 35,000 households (CDC, 2015a). NHANES assesses the health status of approximately 300,000 U.S. adults and children and includes demographic, socioeconomic, dietary, and health-related questions, as well as an examination that includes medical, dental, and physiological measurements and laboratory tests (CDC, 2014, 2015c). NSFG is a continuous survey of men and women age 15 to 49 that assesses family life, marriage and divorce, reproductive health (including pregnancy, infertility, use of contraception), and general health (CDC, 2016). The NSFG sample has ranged from 10,000 to 20,000 (CDC, 2016).
These specific data sources for individual social risk factor indicators and the committee’s recommendations are described in the next chapter.
Adler-Milstein, J., C. M. DesRoches, M. F. Furukawa, C. Worzala, D. Charles, P. Kralovec, S. Stalley, and A. K. Jha. 2014. More than half of US hospitals have at least a basic EHR, but stage 2 criteria remain challenging for most. Health Affairs 33(9):1664-1671.
Adler-Milstein, J., C. M. DesRoches, P. Kralovec, G. Foster, C. Worzala, D. Charles, T. Searcy, and A. K. Jha. 2015. Electronic health record adoption in US hospitals: Progress continues, but challenges persist. Health Affairs 34(12):1-7.
AHRQ (Agency for Healthcare Research and Quality). 2016. CAHPS surveys and guidance. http://www.ahrq.gov/cahps/surveys-guidance/index.html (accessed August 2, 2016).
CDC (Centers for Disease Control and Prevention). 2014. About the National Health and Nutrition Examination Survey. http://www.cdc.gov/nchs/nhanes/about_nhanes.htm (accessed August 11, 2016).
CDC. 2015a. About the National Health Interview Survey. http://www.cdc.gov/nchs/nhis/about_nhis.htm (accessed August 12, 2016).
CDC. 2015b. National Health Interview Survey. https://www.cdc.gov/nchs/data/factsheets/factsheet_nhis.htm (accessed August 4, 2016).
CDC. 2015c. NHANES response rates and population totals. http://www.cdc.gov/nchs/nhanes/response_rates_cps.htm (accessed August 12, 2016).
CDC. 2016. About the National Survey of Family Growth. http://www.cdc.gov/nchs/nsfg/about_nsfg.htm (accessed August 11, 2016).
CMS (Centers for Medicare & Medicaid Services). 2012. Medicare and Medicaid programs: Electronic health record incentive program—stage 2. Federal Register 77(171):53991.
CMS. 2015. Medicare and Medicaid programs; electronic health record incentive program—stage 3 and modifications to meaningful use in 2015 through 2017; Final Rule. Federal Register 80(200):62761-62955.
CMS. 2016a. Consumer Assessment of Healthcare Providers and Systems (CAHPS). https://www.cms.gov/Research-Statistics-Data-andSystems/Research/CAHPS/index.html (accessed August 2, 2016).
CMS. 2016b. EHR incentive programs for eligible professionals: What you need to know for 2016 tipsheet. https://www.cms.gov/Regulations-andGuidance/Legislation/EHRIncentivePrograms/Downloads/2016_EPWhatYouNeedtoKnowfor2016.pdf (accessed August 2, 2016).
CMS. 2016c. Limited data set files: Medicare Current Beneficiary Survey (MCBS). https://www.cms.gov/Research-Statistics-Data-and-Systems/Files-forOrder/LimitedDataSets/MCBS.html (accessed August 2, 2016).
CMS. 2016d. Medicare Current Beneficiary Survey (MCBS). https://www.cms.gov/ResearchStatistics-Data-and-Systems/Research/MCBS/index.html?redirect=/mcbs (accessed August 2, 2016).
DesRoches, C. M., C. Worzala, M. S. Joshi, P. D. Kralovec, and A. K. Jha. 2012. Small, nonteaching, and rural hospitals continue to be slow in adopting electronic health record systems. Health Affairs 31(5):1-7.
DesRoches, C. M., D. Charles, M. F. Furukawa, M. S. Joshi, P. Kralovec, F. Mostashari, C. Worzala, and A. K. Jha. 2013a. Adoption of electronic health records grows rapidly, but fewer than half of US hospitals had at least a basic system in 2012. Health Affairs 32(8):1478-1485.
DesRoches, C. M., C. Worzala, and S. Bates. 2013b. Some hospitals are falling behind in meeting “meaningful use” criteria and could be vulnerable to penalties in 2015. Health Affairs 32(8):1355-1360.
Gottlieb, L. M., K. J. Tirozzi, R. Manchanda, A. R. Burns, and M. T. Sandel. 2015. Moving electronic medical records upstream: Incorporating social determinants of health. American Journal of Preventative Medicine 48(2):215-218.
Haffer, S. C., and S. E. Bowen. 2004. Measuring and improving health outcomes in Medicare: The Medicare HOS program. Healthcare Financing Review 25(4):1-3.
IOM (Institute of Medicine). 2012. Health IT and patient safety: Building safer systems for better care. Washington, DC: The National Academies Press.
IOM. 2014. Capturing social and behavioral domains and measures in electronic health records: Phase 2. Washington, DC: The National Academies Press
Lawson, E. H., R. Carreón, G. Veselovskiy, and J. J. Escarce. 2011. Collection of language data and services provided by health plans. American Journal of Managed Care 17(12):e479-e487.
McNabb, J., D. Timmons, J. Song, and C. Puckett. 2009. Uses of administrative data at the Social Security Administration. Social Security Bulletin 69(1).
Medicare Health Outcomes Survey. 2016. What’s new on the Medicare HOS. http://www.hosonline.org/en (accessed August 2, 2016).
NASEM (The National Academies of Sciences, Engineering, and Medicine). 2016. Accounting for social risk factors in Medicare payment: Criteria, factors, and methods. Washington, DC: The National Academies Press.
Nerenz, D. R., R. Carreon, and G. Veselovskiy. 2013a. Race, ethnicity, and language data collection by health plans: Findings from 2010 America’s Health Insurance Plans Foundation-Robert Wood Johnson Foundation survey. Journal of Health Care for the Poor and Underserved 24(4):1769-1783.
Nerenz, D. R., G. M. Veselovskiy, and R. Carreon. 2013b. Collection of data on race/ethnicity and language proficiency of providers. American Journal of Managed Care 19(12):e408-e414.
NHATS (National Health & Aging Trends Study). 2016. NHATS at a glance. http://www.nhats.org (accessed August 4, 2016).
NIA (National Institute on Aging), NIH (National Institutes of Health), and HHS (Department of Health and Human Services). 2007. Growing older in America: The Health and Retirement Study. http://hrsonline.isr.umich.edu/sitedocs/databook/HRS_Text_WEB_Ch3.pdf (accessed September 19, 2016).
ONC (Office of the National Coordinator for Health Information Technology). 2013. What does “interoperability” mean and why is it important?https://www.healthit.gov/providersprofessionals/faqs/what-does-interoperability-mean-and-why-it-important (accessed August 4, 2016).
ONC. 2014a. Health information exchange (HIE). https://www.healthit.gov/HIE (accessed August 4, 2016).
ONC. 2014b. How to implement EHRs, step 5: Achieve meaningful use. https://www.healthit.gov/providers-professionals/ehr-implementation-steps/step-5achieve-meaningful-use (accessed August 4, 2016).
ONC. 2016a. Non-federal acute care hospital electronic health record adoption. http://dashboard.healthit.gov/quickstats/pages/FIG-Hospital-EHR-Adoption.php (accessed September 6, 2016).
ONC. 2016b. Percent of hospitals, by type, that possess certified health IT. http://dashboard.healthit.gov/quickstats/pages/certified-electronic-health-recordtechnology-in-hospitals.php (accessed September 9, 2016).
ONC. 2016c. Percent of REC enrolled physicians by speciality live on an EHR and demonstrating meaningful use. http://dashboard.healthit.gov/quickstats/pages/FIG-RECPhysicians-Live-MU-Speciality.php (accessed September 6, 2016).
ONC. 2016d. Percent of REC enrolled primary care providers by credentials live on an EHR and demonstrating meaningful use. http://dashboard.healthit.gov/quickstats/pages/FIGREC-Primary-Care-Providers-Live-MU-Credentials.php (accessed September 9, 2016).
ONC. 2016e. Quick stats. http://dashboard.healthit.gov/quickstats/quickstats.php (accessed August 4, 2016).
ONC. n.d. CHP participant community: All Chicago making homelessness history. https://www.healthit.gov/sites/default/files/all_chicago_making_homelessness_historyfinal.pdf (accessed August 15, 2016).
ResDAC (Research Data Assistance Center). 2016a. CMS data: Beneficiary enrollment and demographics. https://www.resdac.org/cmsdata/search?f=im_field_data_file_category%3A237 (accessed August 2, 2016).
ResDAC. 2016b. CMS data: Beneficiary surveys. https://www.resdac.org/cmsdata/search?f=im_field_data_file_category%3A45 (accessed August 2, 2016).
ResDAC. n.d.-a. Health and Retirement Survey—Medicare linked data. https://www.resdac.org/cms-data/files/hrs-medicare (accessed August 4, 2016).
ResDAC. n.d.-b. National Health and Aging Trends Study (NHATS)—Medicare linked data. https://www.resdac.org/cms-data/files/nhats (accessed August 4, 2016).
Shields, A. E., P. Shin, M. G. Leu, D. E. Levy, R. M. Betancourt, D. Hawkins, and M. Proser. 2007. Adoption of health information technology in community health centers: Results of a national survey. Health Affairs 26(5):1373-1383.
Sonnega, A., J. D. Faul, M. B. Ofstedal, K. M. Langa, J. W. Phillips, and D. R. Weir. 2014. Cohort profile: The Health and Retirement Study (HRS). International Journal of Epidemiology 43(2):576-585.
U.S. Census Bureau. 2013. American Community Survey information guide. https://www.census.gov/content/dam/Census/programssurveys/acs/about/ACS_Information_Guide.pdf (accessed August 11, 2016).
U.S. Census Bureau. 2015. American Community Survey: Sample size definitions. http://www.census.gov/programs-surveys/acs/methodology/sample-size-and-dataquality/sample-size-definitions.html (accessed August 12, 2016).
U.S. Census Bureau. 2016. American Community Survey: Areas published. http://www.census.gov/programs-surveys/acs/geography-acs/areas-published.html (accessed August 12, 2016).